The Burke Medical Research Institute, Department of ... publications/2013Haskew.pdf · amp and...

*The Burke Medical Research Institute, Department of Neurology and Neuroscience, Weill Medical

College of Cornell University, White Plains, New York, USA

†Neural Regeneration Laboratory and Ottawa Institute of Systems Biology and Department of

Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, Canada

AbstractAstrocytes are critical for the antioxidant support of neurons.Recently, we demonstrated that low level hydrogen peroxide(H2O2) facilitates astrocyte-dependent neuroprotection inde-pendent of the antioxidant transcription factor Nrf2, leaving theidentity of the endogenous astrocytic Nrf2 activator to ques-tion. In this study, we show that an endogenous electrophile,15-deoxy-D12,14-prostaglandin J2 (15d-PGJ2), non-cellautonomously protects neurons from death induced by deple-tion of the major antioxidant glutathione. Nrf2 knockdown inastrocytes abrogated 15d-PGJ2’s neuroprotective effect aswell as 15d-PGJ2 facilitated Nrf2-target gene induction. Incontrast, knockdown of the transcription factor peroxisomeproliferator activated-receptor gamma (PPARc), a well-char-acterized 15d-PGJ2 target, did not alter 15d-PGJ2 non-cellautonomous neuroprotection. In addition, several PPARc

agonists of the thiazolidinedione (TZD) family failed to induceneuroprotection. Unexpectedly, however, the TZD troglitazone(which contains a chromanol moiety found on vitamin E)induced astrocyte-mediated neuroprotection, an effect whichwas mimicked by the vitamin E analogs alpha-tocopherol oralpha-tocotrienol. Our findings lead to two important conclu-sions: (i) 15d-PGJ2 induces astrocyte-mediated neuroprotec-tion via an Nrf2 but not PPARc mediated pathway, suggestingthat 15d-PGJ2 is a candidate endogenous modulator of Nrf2protective pathways in astrocytes; (ii) selective astrocytetreatment with analogs or compounds containing the chrom-anol moiety of vitamin E facilitates non-cell autonomousneuroprotection.Keywords: astrocytes, neuroprotection, Nrf2, PPARc,vitamin E.J. Neurochem. (2013) 10.1111/jnc.12107

Astrocytes play an integral role in the CNS by regulatingblood flow in response to neuronal activity, modulatingsynaptogenesis and synaptic activity, and providing meta-bolic intermediates to neurons (Dringen et al. 2000; Barres2008; Sofroniew and Vinters 2010). In stroke and other CNSdisorders, astrocytes contribute to neuroprotection in part byproviding antioxidant support, removing excess extracellularlevels of the excitotoxic neurotransmitter glutamate andproviding a barrier to inflammation (Sofroniew and Vinters2010). However, astrocytes may also play a detrimental roleby contributing to cytokine and chemokine release andforming a barrier to neuronal regeneration (Sofroniew andVinters 2010). Thus, exploring therapeutic strategies aimed

Received August 17, 2012; revised manuscript received November 23,2012; accepted November 27, 2012.Address correspondence and reprint requests to Rajiv R. Ratan or

Ren�ee E. Haskew-Layton, Burke Medical Research Institute, Depart-ment of Neurology and Neuroscience, Weill Medical College of CornellUniversity, 785 Mamaroneck Ave, White Plains, NY 10605, USA.E-mails: [email protected], [email protected], [email protected], [email protected] used: 15d-PGJ2, 15-deoxy-D12,14-prostaglandin J2;

ARE, antioxidant response element; CNS, central nervous system; COX,cyclooxygenase; H2O2, hydrogen peroxide; HMOX or HO-1, hemeoxygenase-1; L-PGDS, lipocalin prostaglandin D2 synthase; Nrf2,nuclear factor erythroid 2 related factor 2; PPARc, peroxisomeproliferator activated-receptor gamma; siRNA, short interfering RNA;TZD, thiazolidinedione.

Journal of Neurochemistry © 2012 International Society for Neurochemistry, J. Neurochem. (2013) 10.1111/jnc.12107 1© 2012 The Authors

JOURNAL OF NEUROCHEMISTRY | 2013 doi: 10.1111/jnc.12107

at enhancing the prosurvival and diminishing the prodeathroles that astrocytes play in the CNS are of paramountimportance.Our previous study uncovered the astrocyte-specific neu-

roprotective properties of enzymatically generated H2O2

(Haskew-Layton et al. 2010). Although H2O2 is postulatedto exert some of its biological actions via stabilization of theantioxidant transcription factor nuclear factor erythroid 2related factor 2 (Nrf2) (Bell et al. 2011; Haskew-Laytonet al. 2011), our results showed that the non-cell autonomousneuroprotective effect of astrocytic H2O2 is independent ofNrf2 (Haskew-Layton et al. 2010). In this study we exploredadditional adaptive responses in astrocytes involved inmediating neuroprotection. Specifically, we aimed todelineate the non-cell autonomous neuroprotective effect of15-deoxy-D12,14-Prostaglandin J2 (15d-PGJ2), the levels ofwhich are augmented in response to inflammatory stress(Gilroy et al. 1999).15d-PGJ2 is an electrophilic cyclopentene prostaglandin

produced downstream from cyclooxygenase (COX) andprostaglandin D2 synthase activity (PGD2), which yields15d-PGJ2 following a series of non-enzymatic dehydrationreactions (Gilroy et al. 1999; Scher and Pillinger 2005).Through its direct or indirect interaction with a host ofcellular targets, 15d-PGJ2 exerts multiple biological effectson numerous biological systems, culminating in protectionagainst vascular damage, pulmonary damage, gastrointestinalinjury, myocardial ischemia, pancreatitis, rheumatoid arthri-tis, and numerous other inflammatory injuries (Surh et al.2011). In the CNS, 15d-PGJ2 also exerts a multitude ofbiological effects including anti-inflammatory responses,neurite outgrowth, growth factor secretion, reduction ininfarct in ischemia models, and neurological improvement inmultiple sclerosis (MS) and spinal cord injury (SCI) (Parket al. 2003; Giri et al. 2004; Toyomoto et al. 2004; Pereiraet al. 2006; Zhao et al. 2006; Hatanaka et al. 2010). 15d-PGJ2 has been extensively characterized as an activator ofperoxisome proliferator-activated receptor gamma (PPARc)(Forman et al. 1995; Kliewer et al. 1995). In addition toPPARc, 15d-PGJ2 forms covalent adducts with numerousother cellular targets via Michael addition including Keap1(Nrf2 repressor protein), IjB kinase, thioredoxin, HDACs,UCH-L1, and Ras (Rossi et al. 2000; Shibata et al. 2003;Renedo et al. 2007; Oh et al. 2008; Doyle and Fitzpatrick2010; Koharudin et al. 2010).To explore the mechanism responsible for 15d-PGJ2’s

astrocyte-dependent neuroprotective effect, we focused ontwo transcription factors, PPARc and Nrf2 (Forman et al.1995; Kliewer et al. 1995; Oh et al. 2008; Kobayashi et al.2009). Nrf2 is a potent mediator of astrocyte-dependentneuroprotection in in vitro models of oxidative stress andseveral in vivo models including amyotropic lateral sclerosis(ALS), Parkinson’s disease (PD), and Huntington disease(HD) (Shih et al. 2003; Vargas and Johnson 2009). Nrf2

binds to genes containing the antioxidant response element(ARE) in their promoter regions and enhances the transcrip-tion of antioxidant genes (i.e. heme oxygenase-1, GSHsynthetase) and phase II detoxification genes (i.e. glutathiones-transferase) (Shih et al. 2003). Although important forastrocytic neuroprotection, endogenous activators of Nrf2remain elusive. As 15d-PGJ2 is an electrophile that stabilizesNrf2 via its interaction with Keap1 (Kaspar et al. 2009), 15d-PGJ2 is a likely endogenous signal responsible for activatingNrf2 in brain cells.PPARc is a nuclear receptor transcription factor, which in

addition to exerting antioxidant and anti-inflammatoryresponses, is involved in facilitating adipogenesis, glucosehomeostasis, cell differentiation, and apoptosis (Tontonozand Spiegelman 2008). Similar to Nrf2, PPARc facilitatesneuroprotection in models of stroke, Alzheimer disease(AD), HD, PD, MS, and SCI, via anti-inflammatory orantioxidant-dependent mechanisms (Kapadia et al. 2008;Kiaei 2008). In astrocyte cultures, PPARc dampens inflam-mation in astrocytes and protects neurons from OGD via anastrocyte-dependent mechanism (Tjalkens et al. 2008).Many studies exploring the biological role of PPARc haveutilized compounds from the thiazolidinedione (TZD) fam-ily, yet the specificity of these compounds for PPARc isquestioned by several studies (Davies et al. 2001; Feinsteinet al. 2005; Phulwani et al. 2006). In addition to determiningthe role of PPARc in mediating the astrocyte-dependent 15d-PGJ2 neuroprotective effect, we also sought to establish ifPPARc agonists from the TZD family induce astrocyte-dependent neuroprotection and if so, is their effect dependenton PPARc.In this study we show that 15d-PGJ2 or the TZD

troglitazone potently protect neurons via astrocyte-specificmechanisms independent of PPARc. Our results show that15d-PGJ2 works entirely via Nrf2, and troglitazone likely viaa mechanism involving its vitamin E moiety.

Methods

Additional information regarding the materials and methods can befound in the on-line supplemental material.

Cell culture

All animal procedures were performed according to protocolsapproved by the Institutional Animal Care and Use Committee ofthe Weill Medical College of Cornell University. Primary astrocyteand neuronal cultures were derived from rats obtained from HarlanLaboratories and were prepared as described in (Haskew-Laytonet al. 2010).

Glutathione measurements

Intracellular GSHmeasurements of astrocytes were determined usingthe GSH-Gloe Glutathione Assay kit from Promega (Madison, WI)according to the manufacturer’s protocol. Dithiothreitol was used toconvert GSSG to GSH for total glutathione measurements.

Journal of Neurochemistry © 2012 International Society for Neurochemistry, J. Neurochem. (2013) 10.1111/jnc.12107© 2012 The Authors

2 R. E. Haskew-Layton et al.

Neuronal viability

Neuronal viability was quantified using the MAP2 quantificationmethod developed by Carrier et al. (Carrier et al. 2006), as modifiedin Haskew-Layton et al. (Haskew-Layton et al. 2010).

Phase contrast imaging

Phase contrast images were taken with a Nikon Eclipse TS 100(Nikon Instruments Inc, Melville, NY) using a 40x objective.

15d-PGJ2 quantitation

Total lipids were extracted using a modified Bligh and Dyer method(Ryan et al. 2009). 15d-PGJ2 levels were analyzed using a15-deoxy-D12,14-PGJ2 EIA kit (Enzo Life Sciences, Lörrach,Germany) as per manufacturer’s protocol with modifications.

Quantitative PCR

Total RNA was isolated from primary rat astrocytes using theNucleoSpin® RNA II kit (Clontech, Mountain View, CA). ATaqMan® RNA-to-CTe 1-Step Kit (Applied Biosystems, Carlsbad,CA) was used for both the reverse transcription reaction and thereal-time polymerase chain reaction (PCR), according to themanufacturer’s protocol.

Luciferase reporter assays

PPRE-luciferase: astrocytes were cotransfected with the PPRE X3-TK-luc plasmid upstream of a luciferase reporter [Addgene Plasmid1015 (Kim et al. 1998)] and TK renilla plasmid. Luciferase activitywas measured using a Dual-Luciferase assay kit (Promega). ARE-luciferase: astrocytes were transduced as described in Haskew-Layton et al. (Haskew-Layton et al. 2010) with an adenoviruscontaining the ARE derived from the human NADPH quinoneoxidoreductase gene, upstream of a luciferase reporter (Moehlenk-amp and Johnson 1999). Luciferase activity was measured using aluciferase assay substrate reaction buffer (Promega) containingluciferin.

Western blots

Following standard sodium dodecyl sulfate–polyacrylamide gelelectrophoresis and transfer procedures, proteins were incubatedwith primary antibodies [1 : 1000 anti-HMOX (Enzo Life Sciences,Farmingdale, NY); anti-actin 1 : 5000 (Sigma, St. Louis, MO)],followed by secondary antibody incubation (1 : 20 000). Theimmunoblot bands were detected using the Odyssey infraredimaging system (LI-COR Biosciences, Lincoln, NE). Images areshown in gray scale for clarity.

siRNA transfections

Knockdown of PPARc and Nrf2 was achieved using shortinterfering RNA (siRNA). Commercially available rat PPARcsiRNA duplexes were from SABiosciences (Valencia, CA). RatNrf2 siRNAs were custom designed according to the codingsequences described in Villacorta et al. (2007) and Haskew-Laytonet al., (2010) and synthesized by Qiagen (Valencia, CA).

Statistics

Statistics were performed with Statistica software, StatSoft, Tulsa,OK; or Prism v.6.0 (GraphPad Software, La Jolla, CA); data are theaverages � SEM for three to five experiments per group. Data

were analyzed using one-way ANOVA followed by post hocDunnett’s (for multiple comparisons) or unpaired t-tests (whencomparing two groups).

Results

In our previous study, we characterized a glutathionedepletion-induced model of oxidative stress in a neuron-astrocyte coculture model (Haskew-Layton et al. 2010).In this model the glutamate analog homocysteic acid (HCA)inhibits the xC

� transporter complex to block cystine uptake,subsequently depleting the glutathione precursor cysteineand leading to substantial glutathione depletion in bothastrocytes and immature neurons, but only death in theimmature neurons (Haskew-Layton et al. 2010) (Fig. 1a andb). As immature neurons lack NMDA receptors, we verifiedin our previous study that that HCA-induced death ofneurons in the cocultures is mediated by oxidative stress-independent of excitotoxicity (Haskew-Layton et al. 2010).To test the role of 15d-PGJ2 in protecting neurons in anastrocyte-neuron coculture, immature neurons were plated inthe presence of 15d-PGJ2 on top of a monolayer of confluentastrocytes followed by HCA treatment. Indeed, 15d-PGJ2potently protected the cocultured neurons from HCA-induced death (Fig. 1c and e). To determine if the mecha-nism is astrocyte-specific, astrocytes were pre-treated with15d-PGJ2 prior to neuronal plating in the presence of HCA.In support of an astrocyte-specific role, astrocytic 15d-PGJ2pre-treatment protected the cocultured neurons despite thefact that the neurons were not directly exposed to 15d-PGJ2(Fig. 1d and f).To verify the role of endogenously produced 15d-PGJ2 in

mediating astrocytic neuroprotection, astrocyte-neuron co-cultures were treated with the lipocalin prostaglandin D2synthase (L-PGDS) inhibitor, AT-56. L-PGDS catalyzes theconversion of PGH2 to PGD2, which through a series ofnon-enzymatic dehydration reactions yields 15d-PGJ2 (Surhet al. 2011). Indeed, cocultures treated with AT-56 showedenhanced neuronal death in response to AT-56 (Fig. 1g),which correlated with attenuated 15d-PGJ2 levels in astro-cytes (Fig. 1h). In support of an astrocyte-dependent mech-anism in mediating endogenous 15d-PGJ2’s neuroprotection,neuronal enriched cultures treated with AT-56 were protectedfrom HCA-induced death (Figure S1), which is in contrastwith the accelerated neuronal death observed in the AT-56treated cocultures.To determine if PPARc is important for the 15d-PGJ2

response, astrocytes were pre-treated with PPARc agonists ofthe TZD family prior to neuronal plating with HCA.Although the TZD troglitazone mimicked the astrocyte-dependent neuroprotective response of 15d-PGJ2 (Fig. 2a–c,and g), the other TZDs cigltazone, rosiglitazone or pioglit-azone failed to induce astrocytic neuroprotection (Fig. 2d–f,and h–j), casting doubt on the involvement of PPARc.

© 2012 The AuthorsJournal of Neurochemistry © 2012 International Society for Neurochemistry, J. Neurochem. (2013) 10.1111/jnc.12107

Astrocytic 15d-PGJ2 confers neuroprotection 3

To completely rule out the possible involvement of PPARcin mediating the astrocyte-dependent neuroprotective effectof 15d-PGJ2, PPARc was knocked down via siRNA(Fig. 3a). To verify the functionality of the knockdown,PPARc activity following siRNA knockdown was measuredusing a peroxisome proliferator response element (PPRE)promoter cloned upstream of a luciferase reporter. Indeed,the PPARc siRNAs abrogated 15d-PGJ2 and ciglitazone-induced PPRE-luciferase activity (Fig. 3b). Of note,troglitazone itself failed to induce a significant increase inPPRE-luciferase activity, providing further support thattroglitazone does not protect via a PPARc-dependent mech-anism (Fig. 3b). As expected from the lack of effectivenessof the structurally diverse TZDs on neuroprotection in ourcoculture oxidative stress model, PPARc knockdown failed

to abrogate 15d-PGJ2’s astrocyte-dependent neuroprotectiveeffect (Fig. 3c).As our data argue against the involvement of PPARc in

15d-PGJ2 protection, we next tested if Nrf2 is necessaryusing a previously validated siRNA approach (Haskew-Layton et al. 2010) (Fig. 4a). Confirming the functionality ofthe Nrf2 knockdown, the Nrf2-siRNAs completely abrogatedthe effects of sulforaphane (a classical Nrf2 activator) and15d-PGJ2 on ARE-luciferase activity (Fig. 4b) and dimin-ished 15d-PGJ2’s enhancement of the Nrf2-regulated proteinheme oxygenase-1 (Fig. 4c). However, although this level ofNrf2 knockdown completely abrogated the astrocytic neuro-protective effect of sulforaphane, the effect of 15d-PGJ2 wasonly partially abolished (Fig. 4d). Thus, to determine if analternative mechanism also contributes to the astrocytic

(a) (b)

(c) (d)

(e) (f)

(g) (h)

Fig. 1 15-Deoxy-D12,14-prostaglandin J2(15d-PGJ2) protects neurons fromoxidative stress via an astrocyte-specific

mechanism. (a and b) Immature neuronswere plated in the absence (a) or presence of(b) the glutamate analog, 5 mMhomocysteic

acid (HCA), on top of a confluent layer ofastrocytes, and imaged 48 h later. (c)Neurons were plated in the presence of3 lM 15d-PGJ2 + 5 mM HCA on top of a

monolayer of astrocytes for 48 h. (d)Astrocytes were pre-treated with 3 lM 15d-PGJ2 for 18 h followed by wash-off and the

plating of neurons on top of the astrocyteswith 5 mM HCA for 48 h. (e) Concentrationresponse curve for 15d-PGJ2 treatment,

added with neurons (+/� HCA) when platedon astrocyte monolayer for 48 h [as seen in(c)]. ***p < 0.001, versus HCA control. (f)

Concentration response curve for 15d-PGJ2pre-treatment in astrocytes (18 h) followedby neuronal plating � 5 mM HCA for 48 h[as seen in (d)]. **p < 0.01 or ***p < 0.001,

versus HCA control. (g) Astrocytes were pre-treated with� AT-56 for 18 h followed by theplating of neurons (on top of the confluent

layer of astrocytes) in the presence of � AT-56 and � HCA for 48 h. *p < 0.05, versusHCA control (h) Astrocytes were treated

with � 150 lM AT-56 for 18 h. *p < 0.05,versus DMSO control.



neuroprotective effect of 15d-PGJ2, or further moleculardeletion is required to completely suppress the effect, Nrf2levels were further depleted (Fig. 4e), leading to completeabrogation of 15d-PGJ2’s effect on astrocyte-inducedneuroprotection (Fig. 4f). As astrocytes can protect neuronsfrom oxidative stress via the up-regulation of glutathionebiosynthesis via an Nrf2-dependent mechanism (Shih et al.2003), we provide further support for the involvement ofNrf2, by showing that 15d-PGJ2 preserves glutathione levelsin astrocytes treated with HCA (Fig. 4g).As our data point to an exclusive Nrf2-dependent

astrocytic neuroprotective mechanism for 15d-PGJ2, wesought to determine if Nrf2 signaling is favored over PPARcsignaling in astrocytes. Thus, we looked at mRNA levels forheme oxygenase-1, which contains both Nrf2 and PPARc-sensitive promoter regions (ARE and PPRE regions,respectively) (Prestera et al. 1995; Kronke et al. 2007),and found that heme oxygenase-1 (HMOX) mRNA inducedby 15d-PGJ2 treatment in astrocytes is regulated by Nrf2 but

not PPARc (Fig. 5a). As well, 15d-PGJ2 enhanced themRNA levels of NADPH quinine oxidoreducatase 1 (con-tains the ARE but not the PPRE promoter regions) (Li andJaiswal 1994) but failed to alter mRNA levels of lipoproteinlipase-1 (LPL), which contains PPRE but not the AREpromoter regions (Schoonjans et al. 1996) (Fig. 5b and c).Our results showing that Nrf2 but not PPARc mediates

the effect of 15d-PGJ2, coupled with the lack of effect ofseveral PPARc agonists of the TZD family, suggest that theTZD troglitazone leads to astrocytic neuroprotection inde-pendent of PPARc. Consistent with this, PPARc knockdowndid not abrogate troglitazone’s protective effect (Fig. 6a). Inaddition, because 15d-PGJ2’s robust effect is mediated byNrf2- and it has been postulated that troglitazone activatesNrf2 (Shih et al. 2012)- we tested for Nrf2’s potentialinvolvement, but found that Nrf2-knockdown did not altertroglitazone-induced astrocytic neuroprotection (Fig. 6b). Asthe effect of troglitazone was in contrast with other TZDs,we looked for structural differences as an explanation of the

(a) (b) (c)

(d)

(g) (h)

(i) (j)

(e) (f)

Fig. 2 The thiazolidinedione (TZD)troglitazone, but not other TZDs, protects

neurons from glutathione depletion-inducedoxidative stress via an astrocyte-specificmechanism. (a and b) Immature neuronswere plated in the absence (a) or presence of

(b) the glutamate analog, 5 mMhomocysteicacid (HCA), on top of a confluent layer ofastrocytes, and imaged 48 h later. (c–f)

Astrocytes were pre-treated for 18 h withthe TZDs troglitazone (c), ciglitazone (d),pioglitazone (e), and rosiglitazone (f).

Following wash-off, immature neurons wereplated in the presenceof 5 mMHCA for 48 h.(g–j) Concentration response curve fortroglitazone (g), ciglitazone (h), pioglitazone

(i), and rosiglitazone (j). Experimentsperformed as described for (c–f) and showresults for neurons plated in the presence or

absence of 5 mM HCA. *p < 0.05, versusHCA control.



divergent effects. Troglitazone is unique in that it containsthe same chromanol ring structure as that of vitamin E (fatsoluble family of a, ß, c, and d forms of tocopherols andtocotrienols), a structure absent from other TZDs. Thus, todetermine if the vitamin E chromanol ring structure could

be responsible for troglitazone’s effect, we tested twobiologically active forms of vitamin E, a-tocopherol anda-tocotrienol, and found that astrocyte pre-treatmentwith these compounds mimicked the astrocyte-dependentneuroprotective effect of troglitazone (Fig. 6c and d).Although vitamin E analogs are well-established neuropro-tective agents, our data present a novel mechanism of actionfor vitamin E involving non-cell autonomous protection ofneurons by astrocytes.

Discussion

In this study, we show that astrocytic 15d-PGJ2 providespotent non-cell autonomous neuroprotection from oxidativestress. Using siRNA-mediated knockdown, we demonstratethat 15d-PGJ2’s effect is entirely mediated by Nrf2 butindependent of PPARc, suggesting that 15d-PGJ2 may bethe endogenous molecule responsible for activating anadaptive antioxidant transcriptional response in astrocytesvia activation of Nrf2. In addition to the potent non-cellautonomous neuroprotection seen with 15d-PGJ2, we alsounexpectedly found that the PPARc agonist of the thiazo-lidinedione family, troglitazone, induces astrocyte-specificneuroprotection independent of PPARc, but likely via amechanism involving its vitamin E chromanol ring structure.Understanding the adaptive responses that astrocytes

undergo in response to acute stress or neurodegenerativeconditions will contribute to the development of noveltherapies. Transcription factor signaling represents a criticalaspect of adaptive cellular events that allow cells to ‘sense’alterations in their environment and in response, facilitatean array of specific gene expression changes. Nrf2 is anintegral transcription factor that facilitates astrocytic neuro-protection by promoting the expression of genes involved inglutathione synthesis and release in astrocytes (Shih et al.2003). Several studies showed that exogenous Nrf2 activa-tors or heterologous astrocytic expression of Nrf2 inducesrobust neuroprotection, while Nrf2 knockout exacerbatesneuronal death (Shih et al. 2003; Vargas and Johnson2009). However, because these studies relied on synthetic ornutriceutical Nrf2 activators, or forced expression of Nrf2,the specific endogenous molecule that spurs adaptive Nrf2-mediated astrocytic neuroprotection remains unknown.Basal expression of Nrf2 in the cytoplasm is maintained at

low levels via the Nrf2-repressor protein Keap1 (Kasparet al. 2009). Several in vitro studies show that specificresidues on Keap1 are targets for nucleophilic attack leadingto Nrf2 stabilization, thus Keap1 is thought to be a criticalsensor for intracellular redox changes (Dinkova-Kostovaet al. 2002; Hong et al. 2005; Kobayashi et al. 2009). Thereactive oxygen species H2O2 was considered a critical signalresponsible for Nrf2 stabilization via Keap1 oxidation (Bellet al. 2011). However, our recent study showed that althoughenzymatic H2O2 generation in astrocytes lead to neuropro-

(a)

(b)

(c)

Fig. 3 Astrocytic 15-deoxy-D12,14-prostaglandin J2 (15d-PGJ2) treat-ment protects cocultured neurons from oxidative stress independent

of peroxisome proliferator activated-receptor gamma (PPARc).(a) Astrocyte cultures were treated with siRNA targeted againstPPARc for ~ 42 h prior to RNA isolation. ***p < 0.001, versus si-

Scramble. (b) Astrocytes were transfected with PPRE-Luciferaseplasmids for 24 h followed by 18 h treatments with 3 lM 15d-PGJ2,300 lM troglitazone, or 300 lM ciglitazone (total siRNA treatment

~42 h). *p < 0.05, versus 15d-PGJ2 control or ciglitazone control. (c)Astrocytes were transfected with PPARc siRNAs for 24 h followed by18 h 15d-PGJ2 treatment (total siRNA treatment ~42 h). Followingwash-off of the 15d-PGJ2 treatment, neurons were plated � 5 mM

homocysteic acid (HCA) for 48 h. **p < 0.01, versus 15d-PGJ2control, 15d-PGJ2 + siScramble, 15d-PGJ2 + siP1, and 15d-PGJ2 + siP2.



tection, this effect was independent of Nrf2 (Haskew-Laytonet al. 2010, 2011), highlighting that both Nrf2-dependentand independent pathways in astrocytes are important forneuroprotection. Thus in our current study, we exploredwhether the astrocyte-dependent neuroprotective effectsdriven by 15d-PGJ2 involved Nrf2 as well as other 15d-PGJ2-targeted pro-survival pathways, namely PPARc.

15d-PGJ2 is a cyclopentenone prostaglandin produced viaa series of dehydration reactions downstream from COX andPGD2 synthase metabolism; as such 15d-PGJ2 levels are up-regulated in response to inflammatory stress (Gilroy et al.1999; Scher and Pillinger 2005). Because of an electrophiliccarbon at an a, b-unsaturated ketone within its cyclopentenering, 15d-PGJ2 has numerous protein targets, forming

(a) (b)

(c)

(d)

(e)

(g)

(f)

Fig. 4 15-Deoxy-D12,14-prostaglandin J2 (15d-PGJ2) confersneuroprotection via an astrocytic specific mechanism involving

nuclear factor erythroid 2 related factor 2 (Nrf2). (a) Astrocytecultures were treated with siRNA targeted against Nrf2 for ~ 42 h.***p < 0.001, versus siScramble. (b) ARE-luciferase transduced

astrocytes were treated with siRNA targeted against Nrf2 for 24 hfollowed by treatment with 3 lM 15d-PGJ2 or 5 lM sulforaphane for18 h (total siRNA treatment �42 h). **p < 0.01, versus sulfora-phane no treatment, ***p < 0.001, versus 15d-PGJ2 no treatment.

(c) Astrocytes were transfected for 24 h with siRNAs for Nrf2followed by 18 h 3 lM 15d-PGJ2 treatment (total siRNA treatment~ 42 h). Astrocyte lysates were immunblotted for heme oxygenase-1

(HO-1) and b-actin levels. Lane 1: recombinant HO-1 protein. Lane2: control lysate. Lanes 3–7: + 15d-PGJ2. Tsx: transfection control.

Scr: siScrambled. Si-1,-2: Nrf2 siRNAs. (d) Astrocytes were trans-fected with Nrf2 siRNAs for 24 h followed by 18 h 3 lM 15d-PGJ2

or 5 lM sulforaphane treatment (total siRNA treatment ~ 42 h).Following wash-off of 15d-PGJ2 or sulforaphane, neurons wereplated � 5 mM homocysteic acid (HCA) for 48 h. *p < 0.05, versus

no drug for each respective group. (e) Astrocytes were treated withsiRNA targeted against Nrf2 for ~ 42 h. ***p < 0.001, versussiScramble. (f) Astrocytes were transfected with Nrf2 siRNAs for24 h (siN1 or siN2) followed by 18 h 3 lM 15d-PGJ2 (total siRNA

treatment ~ 42 h). Following wash-off of 15d-PGJ2, neurons wereplated � 5 mM HCA for 48 h. ***p < 0.001, versus HCA control,*p < 0.05, versus HCA control. (g) Astrocytes were treated with

5 mM HCA � 3 lM 15d-PGJ2 for 48 h. *p < 0.05, versus HCAcontrol.



covalent adducts via Michael addition with specific cysteineresidues on proteins (Rossi et al. 2000; Shibata et al. 2003;Renedo et al. 2007; Doyle and Fitzpatrick 2010; Koharudinet al. 2010). As such, 15d-PGJ2 covalently binds with Keap1leading to Nrf2 stabilization (Yamamoto et al. 2008;Kansanen et al. 2009). In line with this, 15d-PGJ2 has beenfound to exert many of its biological effects via Nrf2activation, such as HMOX gene induction, anti-atheroscle-rotic gene induction, protection against acute lung injury,and suppression of inflammatory pathways (Syed et al.2010; Surh et al. 2011). In the CNS, the protective effectsof 15d-PGJ2 in models of stroke, MS, and SCI have largelybeen attributed to the inactivation of NFkB or activation ofPPARc and the subsequent dampening of pro-inflammatoryresponses (Diab et al. 2002; Pereira et al. 2006; Zhao et al.2006; Genovese et al. 2008; Kerr et al. 2008). In astrocytesspecifically, in addition to an anti-inflammatory rolereported for 15d-PGJ2 (Park et al. 2003; Giri et al. 2004;Phulwani et al. 2006; Zhao et al. 2006), other reportedbiological effects of 15d-PGJ2 include enhanced growthfactor secretion, increased glucose utilization and dimin-ished glutamate transporter expression (Toyomoto et al.2004; Garcia-Bueno et al. 2007); which have also beenlargely attributed to blocking NFkB or enhancing PPARc.Other studies have also suggested the involvement of Nrf2in mediating the protective effects of 15d-PGJ2 in nervoussystem cells, including enteric glia (Kim et al. 2008; Abdoet al. 2012). In this study, we provide strong evidence forthe involvement of Nrf2 by using siRNA-mediated knock-down, demonstrating a critical role of Nrf2 in mediating15d-PGJ2-dependent non-cell autonomous neuroprotection.In vivo levels of 15d-PGJ2 have been reported in the ng/

mL range (Rajakariar et al. 2007), leaving to question

whether the levels of 15d-PGJ2 required to induce a biolog-ical response are relevant. However, because only ~1–4% of15d-PGJ2 added to serum containing medium remains ‘free’to enter cells (Rajakariar et al. 2007; Oh et al. 2008), thelowest ‘effective’ concentration of 15d-PGJ2 required toinduce astrocytic-neuroprotection in this study may be as lowas 10 nM (1% of 1 lM, lowest concentration shown toinduce protection), suggesting a true biological effect of 15d-PGJ2 in astrocytes. Finally, in support of a role ofendogenous 15d-PGJ2 in mediating the astrocytic neuropro-tection, inhibition of L-PGDS, an upstream enzyme whichconverts PGH2 to PGD2 (a precursor to 15d-PGJ2),exacerbates glutathione depletion-induced neuronal death inthe astrocyte-neuron coculture model. In contrast however,L-PGDS inhibition in neurons cultured alone yields protec-tion from glutathione depletion-induced death. These datasuggest that endogenous 15d-PGJ2 is important for astrocyticneuroprotection, but may hinder intrinsic neuroprotectivepathways, an area for future investigation.We initially speculated that in addition to Nrf2, PPARc

played a role in 15d-PGJ2’s astrocyte-dependent neuropro-tective effect because of its established role in neuroprotec-tion. PPARc is a transcription factor which enhances thetranscription of antioxidant genes containing the peroxisomeproliferator response element (PPRE) within their enhancer/promoter region and represses pro-inflammatory genesthrough a transrepression mechanism (Tontonoz and Spieg-elman 2008). As such, others have shown that PPARc agonisttreatment in astrocytes induces increased glucose metabolism,increased glutamate transporter expression, repression ofcytokine production and decreased iNOS levels, andenhanced neuronal survival (Kapadia et al. 2008; Kiaei2008; Tontonoz and Spiegelman 2008). In in vivo models,

(a)(b)

(c)

Fig. 5 15-Deoxy-D12,14-prostaglandin J2 (15d-PGJ2) favorsnuclear factor erythroid 2 related factor 2 (Nrf2) dependentsignaling over peroxisome proliferator activated-receptor gamma

(PPARc) signaling in astrocytes. (a) Astrocytes were transfectedwith siRNAs for Nrf2 (50 nM) or PPARc (50 nM) for 24 h

followed by treatment with 3 lM 15d-PGJ2 for 18 h. ***p < 0.001,versus siScramble, *p < 0.05, versus siScramble. (b and c) Con-centration response of 15d-PGJ2 treatment in astrocytes for 18 h.

*p < 0.05, versus 0 lM 15d-PGJ2, **p < 0.01, versus 0 lM 15d-PGJ2.



PPARc enhances neuroprotection in stroke models, models ofMS, ALS, AD, and HD via both antioxidant and anti-inflammatory mechanisms (Kapadia et al. 2008; Kiaei 2008;Zhao et al. 2009). Because glutathione depletion both inducesoxidative stress in cocultures (as characterized in our previousstudy) (Haskew-Layton et al. 2010) and promotes the releaseof proinflammatory cytokines from astrocytes (Lee et al.2010), we speculated that PPARc activation in astrocytesprotects neurons from glutathione depletion via both antiox-idant and anti-inflammatory actions. However, we found that15d-PGJ2 in astrocytes protected neurons independent ofPPARc and that several PPARc agonists failed to induceastrocytic-neuroprotection. In addition, although others havefound that Nrf2 regulates the expression of PPARc (Cho et al.2010; Shih et al. 2012), and thus Nrf2 and PPARc mediateprosurvival pathways in a cooperative fashion, our datasuggests that in astrocytes PPARc is not critical for Nrf2-

depdendent protection. In addition to showing that PPARc isnot important for astrocytic neuroprotection from oxidativestress, we also show that despite having both ARE andPPRE consensus promoter/enhancer sequences, HMOXgene transcription is not highly regulated by PPARc inastrocytes, but is highly regulated by Nrf2.While exploring the potential involvement of PPARc in

mediating the effect of 15d-PGJ2, we unexpectedly foundthat the PPARc agonist troglitazone, but not other com-pounds from the same TZD family, induced potent astrocyte-specific neuroprotection. We confirmed that troglitazonelikely worked via an off-target effect because PPARcknockdown did not abolish troglitazone’s astrocyte-depen-dent neuroprotection. As well, we found that Nrf2 knock-down also failed to abolish troglitazone’s astrocyticneuroprotective effect, highlighting a unique Nrf2-indepen-dent pathway for astrocytic neuroprotection.

(a) (b)

(c)

(e) (f)

(g) (h)

(d)

Fig. 6 The troglitazone effect isindependent of peroxisome proliferatoractivated-receptor gamma (PPARc) andnuclear factor erythroid 2 related factor 2

(Nrf2), but is likely mediated via a ‘vitamin E’moiety; analogous astrocyte-dependentneuroprotection triggered by a-tocopherol

and a-tocotrienol. (a and b) Astrocytecultures were treated with siRNAs targetedagainst PPARc (siP-1,-2; 50 nM – a) or Nrf2

(siN-1,-2–b) for 24 h. Astrocytes were thentreated with 300 lM troglitazone for 18 hfollowed by wash-off and neuronal plating

� 5 mM homocysteic acid (HCA) for 48 h(total siRNA treatment �42 h). **p < 0.01,versus HCA control, *p < 0.05, versus HCAcontrol. (c and d) Astrocytes were pre-

treated with alpha-tocopherol (c) or alpha-tocotrienol (d) for 18 h followed by wash-offand the plating of neurons� 5 mM HCA for

48 h. **p < 0.01, versus HCA control,***p < 0.001, versus HCA control. (e)Control immature neurons plated over non-

treated astrocytes for 48 h. (f) Immatureneurons were plated with 5 mMHCA on topof confluent astrocytes for 48 h. (g and h)Astrocytes were pre-treated with 300 lM

alpha-tocopherol (e) or 300 lM alpha-tocotrienol (f) for 18 h followed by wash-offand the plating of neurons + 5 mM HCA for

48 h.



TZDs are anti-diabetic drugs thought to work via aPPARc-dependent mechanism to affect lipid metabolism(Tontonoz and Spiegelman 2008). However, as in this study,the involvement of PPARc in mediating the biologicaleffects of TZDs has been called into question by severalstudies (Davies et al. 2001; Feinstein et al. 2005; Phulwaniet al. 2006). As troglitazone contains the chromanol ringstructure found in vitamin E analogs, a moiety absent fromthe other TZDs, we speculated and confirmed that thevitamin E analogs, namely a-tocopherol and a-tocotrienol,would mimic the effect of troglitazone.Our unique results with troglitazone lead to the novel

finding that the vitamin E analogs a-tocopherol or a-tocotrienol can exert neuroprotection via non-cell autono-mous mechanisms involving astrocytes. Vitamin E, which iscomposed of a, b, c, and d forms of tocopherols andtocotrienols, acts as an antioxidant by scavenging lipidhydroperoxyl radicals (Kaileh and Sen 2010; Sen et al.2010). In addition to direct antioxidant properties, in the CNSvitamin E modulates the activity of several enzymes includingprotein kinase C, src kinases, 5-lipoxygenase, phospholipaseA2, protein phosphatase 2A, and diacylglycerol kinase, aswell as NFkB activation (Gohil et al. 2004; Muller 2010).Thus, not surprisingly neurological deficits are characteristicof vitamin E deficiency related disorders such as abetalipo-proteinemia and ataxia with vitamin E deficiency (AVED)(Muller 2010). Studies from a-tocopherol transfer proteinknockout mice, show that a-tocopherol regulates expressionof the transcription factor retinoic acid orphan receptor a(ROR-a) and ROR-a downstream target genes, including theastrocyte specific gene GFAP (Gohil et al. 2004). In ourmodel, astrocytes are pre-treated with the vitamin E analogsprior to both neuronal plating and glutamate depletion-induced oxidative stress; thus it seems plausible that, ratherthan exerting neuroprotection via direct antioxidant actions,the vitamin E analogs may be promoting neuroprotection viaan astrocyte-specific cell signaling mechanism.In conclusion, our findings show that 15d-PGJ2, troglit-

azone, and vitamin E analogs protect neurons from oxidativestress non-cell autonomously via astrocyte-specific mecha-nisms. We found that, although characterized as PPARcagonists, 15d-PGJ2 and troglitazone work independent ofPPARc to induce astrocytic neuroprotection. In fact, weshow that astrocytic Nrf2 entirely mediates the neuroprotec-tive effect of 15d-PGJ2. On the other hand, troglitazoneworks independent of Nrf2 but may exert its astrocyte-dependent neurprotection via its vitamin E chromanol ringstructure. Identifying the mechanism responsible for troglit-azone and vitamin E’s astrocyte-dependent effects shouldyield novel therapeutic strategies aimed at enhancing astro-cytic neuroprotection. Altogether our findings point toseveral potent astrocyte-dependent neuroprotective path-ways. We suggest that 15d-PGJ2 represents an adaptivepathway in astrocytes whereby increases in 15d-PGJ2 levels,

in response to cell stress, may produce a feedback mecha-nism to counteract oxidative stress via activation of theantioxidant transcription factor Nrf2. We also suggest thatour findings with troglitazone highlight important Nrf2-independent astrocytic neuroprotective pathways and that thepotent neuroprotective actions of vitamin E in vivo mayoccur through both cell autonomous and non-cell autono-mous mechanisms.

Acknowledgements

This study was supported by the Miriam and Sheldon G. Adelsonfoundation and the Hartman foundation. We thank Dr. AlexanderMongin for critical reading of the manuscript; Li Xia, KathrynBurns, and Mark Rivieccio for primary neurons; Natalya Smirnova,Lawrence Wu, Manuela Basso, and Thong Ma for technical help;Dr. Bruce Spiegelman for providing the PPRE-Luc plasmid; and Dr.Jeff Johnson for providing the ARE-Luc plasmid. Authorship credit:conception and design of project, RH-L and RRR; acquisition ofdata, RH-L, JBP, HX, SALB; analysis and interpretation of data,RH-L, SALB, RRR; drafting and revising of manuscript, RH-L andRRR. The authors report no conflicts of interest.

Supporting information

Additional supporting information may be found in the onlineversion of this article at the publisher’s web-site.

Figure S1. Enriched immature neuronal cultures were plated �5 mM HCA for 18 h in the presence or absence of the L-PGDSinhibitor AT-56.

References

Abdo H., Mahe M. M., Derkinderen P., Bach-Ngohou K., Neunlist M.and Lardeux B. (2012) The omega-6 derivative 15d-PGJ2 isinvolved in neuroprotection by enteric glial cells against oxidativestress. J. Physiol. 590, 2739–2750.

Barres B. A. (2008) The mystery and magic of glia: a perspective ontheir roles in health and disease. Neuron 60, 430–440.

Bell K. F., Al-Mubarak B., Fowler J. H. et al. (2011) Mild oxidativestress activates Nrf2 in astrocytes, which contributes toneuroprotective ischemic preconditioning. Proc. Natl Acad. Sci.USA 108, E1–E2.

Carrier R. L., Ma T. C., Obrietan K. and Hoyt K. R. (2006) A sensitiveand selective assay of neuronal degeneration in cell culture.J. Neurosci. Methods 154, 239–244.

Cho H. Y., Gladwell W., Wang X., Chorley B., Bell D., Reddy S. P. andKleeberger S. R. (2010) Nrf2-regulated PPAR{gamma} expressionis critical to protection against acute lung injury in mice. Am.J. Respir. Crit. Care Med. 182, 170–182.

Davies G. F., Khandelwal R. L., Wu L., Juurlink B. H. and Roesler W. J.(2001) Inhibition of phosphoenolpyruvate carboxykinase (PEPCK)gene expression by troglitazone: a peroxisome proliferator-activated receptor-gamma (PPARgamma)-independent,antioxidant-related mechanism. Biochem. Pharmacol. 62, 1071–1079.

Diab A., Deng C., Smith J. D., Hussain R. Z., Phanavanh B., Lovett-Racke A. E., Drew P. D. and Racke M. K. (2002) Peroxisomeproliferator-activated receptor-gamma agonist 15-deoxy-Delta



(12,14)-prostaglandin J(2) ameliorates experimental autoimmuneencephalomyelitis. J. Immunol. 168, 2508–2515.

Dinkova-KostovaA. T.,HoltzclawW.D., ColeR.N., ItohK.,WakabayashiN., Katoh Y., Yamamoto M. and Talalay P. (2002) Direct evidencethat sulfhydryl groups of Keap1 are the sensors regulating inductionof phase 2 enzymes that protect against carcinogens and oxidants.Proc. Natl Acad. Sci. USA 99, 11908–11913.

Doyle K. and Fitzpatrick F. A. (2010) Redox signaling, alkylation(carbonylation) of conserved cysteines inactivates class I histonedeacetylases 1, 2, and 3 and antagonizes their transcriptionalrepressor function. J. Biol. Chem. 285, 17417–17424.

Dringen R., Gutterer J. M. and Hirrlinger J. (2000) Glutathionemetabolism in brain metabolic interaction between astrocytes andneurons in the defense against reactive oxygen species. Eur. J.Biochem. 267, 4912–4916.

Feinstein D. L., Spagnolo A., Akar C., Weinberg G., Murphy P.,Gavrilyuk V. and Dello Russo C. (2005) Receptor-independentactions of PPAR thiazolidinedione agonists: is mitochondrialfunction the key? Biochem. Pharmacol. 70, 177–188.

Forman B. M., Tontonoz P., Chen J., Brun R. P., Spiegelman B. M. andEvans R. M. (1995) 15-Deoxy-delta 12, 14-prostaglandin J2 is aligand for the adipocyte determination factor PPAR gamma. Cell83, 803–812.

Garcia-Bueno B., Caso J. R., Perez-Nievas B. G., Lorenzo P. and Leza J.C. (2007) Effects of peroxisome proliferator-activated receptorgamma agonists on brain glucose and glutamate transporters afterstress in rats. Neuropsychopharmacology 32, 1251–1260.

Genovese T., Esposito E., Mazzon E., Di Paola R., Muia C., Meli R.,Bramanti P. and Cuzzocrea S. (2008) Effect of cyclopentanoneprostaglandin 15-deoxy-delta12,14PGJ2 on early functionalrecovery from experimental spinal cord injury. Shock 30, 142–152.

Gilroy D. W., Colville-Nash P. R., Willis D., Chivers J., Paul-Clark M.J. and Willoughby D. A. (1999) Inducible cyclooxygenase mayhave anti-inflammatory properties. Nat. Med. 5, 698–701.

Giri S., Rattan R., Singh A. K. and Singh I. (2004) The 15-deoxy-delta12,14-prostaglandin J2 inhibits the inflammatory response inprimary rat astrocytes via down-regulating multiple steps inphosphatidylinositol 3-kinase-Akt-NF-kappaB-p300 pathwayindependent of peroxisome proliferator-activated receptorgamma. J. Immunol. 173, 5196–5208.

Gohil K., Godzdanker R., O’Roark E., Schock B. C., Kaini R. R., PackerL., Cross C. E. and Traber M. G. (2004) Alpha-tocopherol transferprotein deficiency in mice causes multi-organ deregulation of genenetworks and behavioral deficits with age. Ann. N. Y. Acad. Sci.1031, 109–126.

Haskew-Layton R. E., Payappilly J. B., Smirnova N. A. et al. (2010)Controlled enzymatic production of astrocytic hydrogen peroxideprotects neurons from oxidative stress via an Nrf2-independentpathway. Proc. Natl Acad. Sci. USA 107, 17385–17390.

Haskew-Layton R. E., Ma T. C. and Ratan R. R. (2011) Reply to Bell etal.: Nrf2-dependent and -independent mechanisms of astrocyticneuroprotection. Proc. Natl Acad. Sci. USA 108, E3–E4.

Hatanaka M., Shibata N., Shintani N., Haba R., Hayata A., HashimotoH. and Baba A. (2010) 15d-prostaglandin J2 enhancement of nervegrowth factor-induced neurite outgrowth is blocked by thechemoattractant receptor- homologous molecule expressed onT-helper type 2 cells (CRTH2) antagonist CAY10471 in PC12cells. J. Pharmacol. Sci. 113, 89–93.

Hong F., Sekhar K. R., Freeman M. L. and Liebler D. C. (2005) Specificpatterns of electrophile adduction trigger Keap1 ubiquitination andNrf2 activation. J. Biol. Chem. 280, 31768–31775.

Kaileh M. and Sen R. (2010) Role of NF-kappaB in the anti-inflammatoryeffects of tocotrienols. J. Am. Coll. Nutr. 29, 334S–339S.

Kansanen E., Kivela A. M. and Levonen A. L. (2009) Regulation ofNrf2-dependent gene expression by 15-deoxy-Delta12,14-prostaglandin J2. Free Radic. Biol. Med. 47, 1310–1317.

Kapadia R., Yi J. H. and Vemuganti R. (2008) Mechanisms of anti-inflammatory and neuroprotective actions of PPAR-gammaagonists. Front. Biosci. 13, 1813–1826.

Kaspar J. W., Niture S. K. and Jaiswal A. K. (2009) Nrf2:INrf2 (Keap1)signaling in oxidative stress. Free Radic. Biol. Med. 47, 1304–1309.

Kerr B. J., Girolami E. I., Ghasemlou N., Jeong S. Y. and David S.(2008) The protective effects of 15-deoxy-delta-(12,14)-prostaglandin J2 in spinal cord injury. Glia 56, 436–448.

Kiaei M. (2008) Peroxisome proliferator-activated receptor-gamma inamyotrophic lateral sclerosis and Huntington’s disease. PPAR Res.2008, 418765.

Kim J. B., Wright H. M., Wright M. and Spiegelman B. M. (1998)ADD1/SREBP1 activates PPARgamma through the production ofendogenous ligand. Proc. Natl Acad. Sci. USA 95, 4333–4337.

Kim J. W., Li M. H., Jang J. H., Na H. K., Song N. Y., Lee C., JohnsonJ. A. and Surh Y. J. (2008) 15-Deoxy-Delta(12,14)-prostaglandin J(2) rescues PC12 cells from H2O2-induced apoptosis throughNrf2-mediated upregulation of heme oxygenase-1: potential rolesof Akt and ERK1/2. Biochem. Pharmacol. 76, 1577–1589.

Kliewer S. A., Lenhard J. M., Willson T. M., Patel I., Morris D. C. andLehmann J. M. (1995) A prostaglandin J2 metabolite bindsperoxisome proliferator-activated receptor gamma and promotesadipocyte differentiation. Cell 83, 813–819.

Kobayashi M., Li L., Iwamoto N., Nakajima-Takagi Y., Kaneko H.,Nakayama Y., Eguchi M., Wada Y., Kumagai Y. and YamamotoM. (2009) The antioxidant defense system Keap1-Nrf2 comprises amultiple sensing mechanism for responding to a wide range ofchemical compounds. Mol. Cell. Biol. 29, 493–502.

Koharudin L. M., Liu H., Di Maio R., Kodali R. B., Graham S. H. andGronenborn A. M. (2010) Cyclopentenone prostaglandin-inducedunfolding and aggregation of the Parkinson disease-associatedUCH-L1. Proc. Natl Acad. Sci. USA 107, 6835–6840.

Kronke G., Kadl A., Ikonomu E. et al. (2007) Expression of hemeoxygenase-1 in human vascular cells is regulated by peroxisomeproliferator-activated receptors. Arterioscler. Thromb. Vasc. Biol.27, 1276–1282.

Lee M., Cho T., Jantaratnotai N., Wang Y. T., McGeer E. and McGeer P.L. (2010) Depletion of GSH in glial cells induces neurotoxicity:relevance to aging and degenerative neurological diseases. FASEBJ. 24, 2533–2545.

Li Y. and Jaiswal A. K. (1994) Human antioxidant-response-element-mediated regulation of type 1 NAD(P)H:quinone oxidoreductasegene expression. Effect of sulfhydryl modifying agents. Eur. J.Biochem. 226, 31–39.

Moehlenkamp J. D. and Johnson J. A. (1999) Activation of antioxidant/electrophile-responsive elements in IMR-32 human neuroblastomacells. Arch. Biochem. Biophys. 363, 98–106.

Muller D. P. (2010) Vitamin E and neurological function. Mol. Nutr.Food Res. 54, 710–718.

Oh J. Y., Giles N., Landar A. and Darley-Usmar V. (2008) Accumulationof 15-deoxy-delta(12,14)-prostaglandin J2 adduct formation withKeap1 over time: effects on potency for intracellular antioxidantdefence induction. Biochem. J. 411, 297–306.

Park E. J., Park S. Y., Joe E. H. and Jou I. (2003) 15d-PGJ2 androsiglitazone suppress Janus kinase-STAT inflammatory signalingthrough induction of suppressor of cytokine signaling 1 (SOCS1)and SOCS3 in glia. J. Biol. Chem. 278, 14747–14752.

Pereira M. P., Hurtado O., Cardenas A. et al. (2006) Rosiglitazoneand 15-deoxy-Delta12,14-prostaglandin J2 cause potentneuroprotection after experimental stroke through noncompletely



overlapping mechanisms. J. Cereb. Blood Flow Metab. 26, 218–229.

Phulwani N. K., Feinstein D. L., Gavrilyuk V., Akar C. and Kielian T.(2006) 15-deoxy-Delta12,14-prostaglandin J2 (15d-PGJ2) andciglitazone modulate Staphylococcus aureus-dependent astrocyteactivation primarily through a PPAR-gamma-independentpathway. J. Neurochem. 99, 1389–1402.

Prestera T., Talalay P., Alam J., Ahn Y. I., Lee P. J. and Choi A. M.(1995) Parallel induction of heme oxygenase-1 andchemoprotective phase 2 enzymes by electrophiles andantioxidants: regulation by upstream antioxidant-responsiveelements (ARE). Mol. Med. 1, 827–837.

Rajakariar R., Hilliard M., Lawrence T., Trivedi S., Colville-Nash P.,Bellingan G., Fitzgerald D., Yaqoob M. M. and Gilroy D. W.(2007) Hematopoietic prostaglandin D2 synthase controls the onsetand resolution of acute inflammation through PGD2 and15-deoxyDelta12 14 PGJ2. Proc. Natl Acad. Sci. USA 104,20979–20984.

Renedo M., Gayarre J., Garcia-Dominguez C. A., Perez-Rodriguez A.,Prieto A., Canada F. J., Rojas J. M. and Perez-Sala D. (2007)Modification and activation of Ras proteins by electrophilicprostanoids with different structure are site-selective.Biochemistry 46, 6607–6616.

Rossi A., Kapahi P., Natoli G., Takahashi T., Chen Y., Karin M. andSantoro M. G. (2000) Anti-inflammatory cyclopentenoneprostaglandins are direct inhibitors of IkappaB kinase. Nature403, 103–108.

Ryan S. D., Whitehead S. N., Swayne L. A. et al. (2009) Amyloid-beta42 signals tau hyperphosphorylation and compromisesneuronal viability by disrupting alkylacylglycerophosphocholinemetabolism. Proc. Natl Acad. Sci. USA 106, 20936–20941.

Scher J. U. and Pillinger M. H. (2005) 15d-PGJ2: the anti-inflammatoryprostaglandin? Clin. Immunol. 114, 100–109.

Schoonjans K., Peinado-Onsurbe J., Lefebvre A. M., Heyman R. A.,Briggs M., Deeb S., Staels B. and Auwerx J. (1996) PPARalphaand PPARgamma activators direct a distinct tissue-specifictranscriptional response via a PPRE in the lipoprotein lipasegene. EMBO J. 15, 5336–5348.

Sen C. K., Rink C. and Khanna S. (2010) Palm oil-derived naturalvitamin E alpha-tocotrienol in brain health and disease. J. Am. Coll.Nutr. 29, 314S–323S.

Shibata T., YamadaT., Ishii T., Kumazawa S., NakamuraH.,Masutani H.,Yodoi J. and Uchida K. (2003) Thioredoxin as a molecular target ofcyclopentenone prostaglandins. J. Biol. Chem. 278, 26046–26054.

Shih A. Y., Johnson D. A., Wong G., Kraft A. D., Jiang L., Erb H.,Johnson J. A. and Murphy T. H. (2003) Coordinate regulation ofglutathione biosynthesis and release by Nrf2-expressing gliapotently protects neurons from oxidative stress. J. Neurosci. 23,3394–3406.

Shih P. H., Hwang S. L., Yeh C. T. and Yen G. C. (2012) Synergisticeffect of cyanidin and PPAR agonist against nonalcoholicsteatohepatitis-mediated oxidative stress-induced cytotoxicitythrough MAPK and Nrf2 transduction pathways. J. Agric. FoodChem. 60, 2924–2933.

Sofroniew M. V. and Vinters H. V. (2010) Astrocytes: biology andpathology. Acta Neuropathol. 119, 7–35.

Surh Y. J., Na H. K., Park J. M., Lee H. N., Kim W., Yoon I. S. and KimD. D. (2011) 15-Deoxy-Delta(1)(2), (1)(4)-prostaglandin J(2), anelectrophilic lipid mediator of anti-inflammatory and pro-resolvingsignaling. Biochem. Pharmacol. 82, 1335–1351.

Syed M. A., Joo M., Abbas Z., Rodger D., Christman J. W., Mehta D.and Sadikot R. T. (2010) Expression of TREM-1 is inhibitedby PGD2 and PGJ2 in macrophages. Exp. Cell Res. 316,3140–3149.

Tjalkens R. B., Liu X., Mohl B., Wright T., Moreno J. A., Carbone D. L.and Safe S. (2008) The peroxisome proliferator-activated receptor-gamma agonist 1,1-bis(3’-indolyl)-1-(p-trifluoromethylphenyl)methane suppresses manganese-induced production of nitricoxide in astrocytes and inhibits apoptosis in cocultured PC12cells. J. Neurosci. Res. 86, 618–629.

Tontonoz P. and Spiegelman B. M. (2008) Fat and beyond: the diversebiology of PPARgamma. Annu. Rev. Biochem. 77, 289–312.

Toyomoto M., Ohta M., Okumura K., Yano H., Matsumoto K., Inoue S.,Hayashi K. and Ikeda K. (2004) Prostaglandins are powerfulinducers of NGF and BDNF production in mouse astrocytecultures. FEBS Lett. 562, 211–215.

Vargas M. R. and Johnson J. A. (2009) The Nrf2-ARE cytoprotectivepathway in astrocytes. Expert Rev. Mol. Med. 11, e17.

Villacorta L., Zhang J., Garcia-Barrio M. T., Chen X. L., Freeman B.A., Chen Y. E. and Cui T. (2007) Nitro-linoleic acid inhibitsvascular smooth muscle cell proliferation via the Keap1/Nrf2signaling pathway. Am. J. Physiol. Heart Circ. Physiol. 293,H770–H776.

Yamamoto T., Suzuki T., Kobayashi A., Wakabayashi J., Maher J.,Motohashi H. and Yamamoto M. (2008) Physiological significanceof reactive cysteine residues of Keap1 in determining Nrf2 activity.Mol. Cell. Biol. 28, 2758–2770.

Zhao X., Zhang Y., Strong R., Grotta J. C. and Aronowski J. (2006) 15d-Prostaglandin J2 activates peroxisome proliferator-activatedreceptor-gamma, promotes expression of catalase, and reducesinflammation, behavioral dysfunction, and neuronal loss afterintracerebral hemorrhage in rats. J. Cereb. Blood Flow Metab. 26,811–820.

Zhao X., Strong R., Zhang J., Sun G., Tsien J. Z., Cui Z., Grotta J. C.and Aronowski J. (2009) Neuronal PPARgamma deficiencyincreases susceptibility to brain damage after cerebral ischemia.J. Neurosci. 29, 6186–6195.



The Burke Medical Research Institute, Department of ... publications/2013Haskew.pdf · amp and...

Documents

Transcript of The Burke Medical Research Institute, Department of ... publications/2013Haskew.pdf · amp and...